IRCloggy #git 2011-10-15

Logs Search ←Prev date Next date→ Channels Documentation

Provider of IRC logs since 2005.
WARNING: As Freenode became unjoinable and lost all warnings in topics, we cannot log channels on Freenode anymore.

2011-10-15

→ kermit joined00:00
→ whit537 joined00:01
→ bawigga joined00:01
bawigga how do i find out where a branch was branched from?00:01
← mcluskydodallas left00:01
cmn bawigga: gitk or git log --oneline --decorate00:03
→ mcluskydodallas joined00:03
← whit537 left00:03
SethRobertson sitaram: Looks like you already do what the person was wanting (and I thought was useful) WRT being able to use LDAP groups00:03
→ whit537 joined00:03
rawtaz lol this is so confusing :p00:05
i can see why ppl choose mercurial and what not :P00:05
im expecting to see, at some point, why ppl choose git though00:06
→ wilmoore joined00:06
cmn is there anything in particular that could do with an update to documentation?00:06
SethRobertson rawtaz: There are really only a handful of commands you need to know for day to day activity.00:07
rawtaz cmn: nah, i am not that deep into it yet. i am currently reading an answer at stackoverflow, so the git docs can certainly not be blamed00:07
http://stackoverflow.com/questions/5798930/whats-the-difference-between-git-rm-cached-asd-and-git-reset-head-asd00:07
SethRobertson rawtaz: Read !book00:07
jast rawtaz: *automatic message* 'Pro Git' is a complete book about git that is available online for free: http://progit.org/book/00:07
rawtaz could be bad explanations as well00:07
SethRobertson: just trying to figure out the opposite of `git add`, thats why00:08
SethRobertson That is a very technical question which I doubt you are going to encounter in your day to day activities00:08
→ makopo joined00:08
bawigga @cmm, perfect thanks00:08
← makopo left00:08
rawtaz because im trying to add a folder without its files and subdirectories00:08
sitaram SethRobertson: not sure; he said he "gets these permissions from LDAP". Could be either, but it *sounded* like authorisation to me not just authentication00:08
rawtaz but git added the entire folder recursively :)00:08
← bawigga left00:08
← dhruvasagar left00:08
niloc132 rawtaz: git doesnt know directories00:09
it just knows files, so adding a directory means to add the files in it00:09
rawtaz THAT explains things00:09
← dreiss left00:09
→ makopo joined00:09
rawtaz SethRobertson: getting the book, thanks for the pointer :)00:10
niloc132 i'm not a huge fan of that design decision, but it does make some things make more sense00:10
rawtaz niloc132: and all your help too00:10
← whit537 left00:10
rawtaz niloc132: is there anything with it that you think is problematic?00:10
SethRobertson rawtaz: git add `find . -maxdepth 1 -type f`00:10
niloc132 not especially, and i am using it on its own, and in conjunction with svn (via git-svn)00:10
→ osxorgate joined00:11
rawtaz SethRobertson: apparently adding a folder without the files in it isnt applicable :)00:11
→ dhruvasagar joined00:11
SethRobertson Very true00:11
→ _Vi joined00:12
→ univate_ joined00:12
rawtaz niloc132: kinda feels like it takes away a bit of complexity but not supporting directories per se but rather just paths00:12
SethRobertson People often create a .gitignore file in such an "empty" directory and add that00:12
rawtaz SethRobertson: why? becuase they dont know better?00:12
SethRobertson Or have the `make` or similar build/install script take care of it00:12
← osxorgate left00:12
SethRobertson Because the directory is not going to stay empty, so you probably will need to ignore the contents. Creating the .gitignore file with "*" in it will take care of two "problems" at the same time. There are other approaches, of course00:13
_Vi Where is GitSurvey's questions or results? (git.wiki.kernel.org is down, survs.com rejects)00:13
rawtaz SethRobertson: i just dont see why they want to create the folder in the first place. or add it, rather. not sure what to call what they do. could it be because when checking out a revision, they want to have an empty folder created (for example an assets folder that published assets in a web app will be written to)?00:14
SethRobertson _Vi: survs.com works for me. A bit slow, but working.00:14
_Vi SethRobertson, "rejects" means "This survey is currently closed. More information can be found on http://git.wiki.kernel.org/index.php/GitSurvey2011".00:15
SethRobertson _Vi: You cannot answer the survey any more. You can only look at the results00:15
_Vi: https://www.survs.com/results/Q5CA9SKQ/P7DE07F0PL00:16
_Vi SethRobertson, Where are results? (Missed the survey time again).00:16
→ whit537 joined00:16
← caseymcg left00:16
_Vi SethRobertson, Thanks. This links should have been on "This survey is currently" page too.00:16
SethRobertson rawtaz: It typically is done for situations where git is being used to manage non-source code. For instance, in a website deployment, you might want to have certain directories not under the control of git (dynamic content, very large files, etc) but have the directories created to simplify installation.00:17
← maletor left00:18
SethRobertson _Vi: Probably, but it may not have been possible after the survey was closed. There was an announcement on the mailing list00:18
← whit537 left00:19
→ baburdick joined00:19
rawtaz SethRobertson: yeah exactly. i would like that.00:19
SethRobertson And I provided the method people choose00:20
rawtaz yeah00:20
← ddkilzer left00:20
rawtaz is it the recommended one?00:20
you said there were other approaches as well00:20
_Vi Is there a low-traffic Git mailing list?00:20
← martinjlowm left00:20
SethRobertson _Vi: Nope. I filter out certain threads in my mail reader, like PATCHES00:21
_Vi SethRobertson, May be there should be special low-traffic mailing list for changelogs when new version is released and global events like such surveys?00:21
→ caseymcg joined00:22
SethRobertson rawtaz: One: make a directory with .gitignore in it which you add. Two: Have a compilation, installation, distribution script, and/or git hook hook create the directory for you.00:22
rawtaz ok. i will read up on that00:22
← jalama left00:22
SethRobertson _Vi: People usually put ANNOUNCE in the subject. Filter on that.00:22
← brian_g left00:22
SethRobertson rawtaz: Note that git hooks, which might sound ideal, cannot typically be set up automatically for you. It is something that you must install after cloning the git repo.00:23
_Vi (Thinking how to do it with POP3 without causing network traffic)00:23
→ eletuchy joined00:24
SethRobertson Why not route it to some mail system which has the ability to filter for you (like gmail perhaps) and then pop there00:24
→ kmc joined00:26
Mpenz sitaram: hi, how are you. Yeah, I understand. I guess, but I don't know for sure, but it'd be good to be able to just see access requests and their latency etc.00:27
→ brendon joined00:28
Mpenz sitaram: Even if these scripts are not available as a part of the src code, perhaps a plugin or some sorts that has a separate repo for people like me ;)00:31
← towski left00:32
← gusnan left00:34
← Spockz left00:34
→ Daphko joined00:37
← univate_ left00:37
→ univate_ joined00:39
← heppy left00:43
← ivorybishop left00:44
← baburdick left00:47
→ baburdick joined00:48
← pheaver left00:50
→ berserkr joined00:50
← berserkr left00:51
→ ISF joined00:52
→ glennpratt joined00:54
← kenperkins left00:54
← afief left00:55
← dhruvasagar left00:57
← mcluskydodallas left00:57
→ mcluskydodallas joined00:58
→ dfr|mac joined00:59
→ osxorgate joined00:59
→ dhruvasagar joined00:59
← apok left01:00
→ jharrys joined01:00
← xpot-mobile left01:00
← caseymcg left01:01
→ xiangfu joined01:02
← osxorgate left01:02
→ caseymcg joined01:03
← blakefrost left01:03
→ FreezingCold joined01:03
rawtaz this is probably dead stupid, but when i do `git status` i get public/assets/ but no files listed under that, in the Untracked files listing. but there are a bunch of directories and files under that folder01:05
why are they not shown?01:05
any small pointer is enough01:05
← bronson left01:05
← caseymcg left01:06
→ caseymcg joined01:07
→ dibber joined01:07
← Araxia left01:08
cmn rawtaz: that means there is nothing under that directory that's tracked01:08
→ seed419 joined01:08
rawtaz so its just git's way of showing this. makes sense01:09
→ guy_brush joined01:09
seed419 git confuses me greatly01:09
rawtaz in a way one could say that the files under that folder are untracked too, though :(01:09
:)*01:09
→ Jay_Levitt joined01:09
rawtaz seed419: me too, but we'll get through it :P01:09
← palithius left01:10
← caseymcg left01:10
seed419 i forked a repository, how do I get a local folder of it on my machine?01:10
rawtaz seed419: on github?01:11
seed419 yeah01:11
→ caseymcg joined01:11
rawtaz that is explained well by them on that site01:11
go to the Help pages :j01:11
seed419 okay, i'll re-read the help section01:11
im an idiot01:11
← zivester left01:11
cmn seed419: read the !book01:11
jast seed419: *automatic message* 'Pro Git' is a complete book about git that is available online for free: http://progit.org/book/01:11
cmn and the manual pages01:11
rawtaz seed419: http://help.github.com/fork-a-repo/01:12
→ zivester joined01:12
seed419 thanks guys01:12
← ruskie left01:12
→ MUILTFN joined01:13
← caseymcg left01:15
→ caseymcg joined01:15
← ben_alman left01:16
→ ben_alman joined01:16
→ ruskie joined01:18
← baburdick left01:18
→ baburdick joined01:18
← ruskie left01:19
FreezingCold Alright, really stupid thing I need to ask: I'm trying to update linux-next, but it's a different branch right (new day)? So should I just totally delete my current clone and re-pull, or is there a faster way?01:19
← EugeneKay left01:19
← ColKurtz left01:20
→ EugeneKay joined01:20
← Daphko left01:22
cmn FreezingCold: fetch the updates01:23
→ osxorgate joined01:24
FreezingCold Like git pull?01:24
cmn like git fetch01:25
pull is fetch+merge01:25
FreezingCold S01:26
Another stupid question, how can I view the kernel name without compiling?01:27
→ ruskie joined01:28
FreezingCold The kernel didn't change names at all... I don't think it updated01:29
kevlarman why would it?01:29
→ neurodrone joined01:29
FreezingCold from like 20111013 to 2011101401:30
← pdtpatrick left01:30
→ dreiss joined01:30
← jaequery left01:31
← SkyFlyer left01:31
kevlarman FreezingCold: fetch doesn't touch your working copy, it's up to you to update that01:31
FreezingCold Yeah that's what I thought, I want to update to the latest one01:32
git pull doesn't work cause I have a bunch of files it complains about01:32
← neurodrone left01:32
kadoban define "complains about"01:34
CareBear\ unstaged changes perhaps01:36
→ soreau joined01:38
soreau What is with these '...skipping...'messages in git log?01:38
→ jalama joined01:39
soreau It's '...skipping...' some important parts I need to see01:39
→ pheaver joined01:41
← jharrys left01:41
soreau This is really annoying01:41
offby1 never seen those01:42
← Vampire0_ left01:42
FreezingCold Well git pull 2>&1 | wgetpaste is taking FOREVER01:42
it basically the following files would have to be removed or something like that to merge01:43
*basically said01:43
← pheaver left01:43
offby1 soreau: no offense, but are you sure that '...skipping...' is coming from the "git log" command, and isn't part of some log messages?01:45
→ johnkpaul joined01:46
soreau offby1: Yes, it is definitely git log.. I could show you an excerpt01:46
The cgit site is fine01:46
← FreezingCold left01:47
→ ZapZ joined01:47
kadoban i know you can configure git log to skip certain things, like merges i think? but i think it just doesn't show them, i don't remember there being a "skipping" message01:47
offby1 me neither01:47
← L0cky left01:48
← baburdick left01:48
offby1 the string "skipping" doesn't appear many places in the git sources, and in no place is it surrounded by ellipses01:49
ajsharp → ajsharp_zz01:49
kevlarman that's what less outputs when you try to scroll more than one page in between updates01:49
soreau This says "...skipping..." http://pastebin.com/wMATMMtt01:50
→ baburdick joined01:50
cmn soreau: that's a message from your pager01:50
soreau cmn: ah01:51
How to 'fix' it?01:51
soreau doesn't know about said pagers01:51
→ khaije|minder joined01:51
kevlarman soreau: don't try to scroll multiple pages in between updates01:51
or ignore it01:51
soreau kevlarman: I don't know what you mean..01:52
Ah yes.. if I scroll, it gives the message. When I use arrow keys, it does not01:52
cmn right, because it's skipping stuff, Ctrl+L should refresh your screen01:53
→ Araxia joined01:54
khaije|minder where should i store my bisect tests? i started out keeping them in the same repo they are run against, but then i realized that would mean they weren't available as HEAD flies around... any suggestions/best-practices?01:54
← johnkpaul left01:54
kevlarman khaije|minder: why do you need to store them?01:54
soreau cmn: kevlarman: thanks01:55
← osxorgate left01:55
khaije|minder kevlarman: i just like to track complexity over time01:55
it's not a requirement but i do need some kind of sensible place to put them01:56
→ joshwa joined01:56
← warthog9 left01:57
offby1 khaije|minder: another repository :)01:59
seriously, I never keep mine01:59
← soreau left01:59
→ L0cky joined02:00
khaije|minder offby1: oh ya? I'm new to the concept and it seems soo incredibly useful that I wanted to track my lessons02:01
cmn khaije|minder: you can just reference the commit that introduced the regression in the commit message of the fix02:02
→ pdtpatrick joined02:02
joshwa can anyone help me figure out why git-svn is committing to the wrong branch? http://pastie.org/269830302:03
← sbell left02:04
→ dexter_e joined02:04
khaije|minder cmn: thats a good idea, though in this case i was talking about the lessons learned while putting together different types of "git bisect"-friendly tests02:04
← bartek left02:05
← mcluskydodallas left02:05
← icwiener left02:05
→ warthog9 joined02:05
→ pheaver joined02:05
← jlatt left02:06
← dr_lepper left02:06
← caseymcg left02:06
offby1 joshwa: sorry, I never understodd that part of git-svn02:07
*understood02:07
joshwa i had it working before, but my company changed the project layout and now it's all b0rked02:07
← jalama left02:07
joshwa rebase does the same thing02:08
→ dankest joined02:08
→ shiba_yu36 joined02:08
← guy_brush left02:12
→ Vampire0 joined02:12
→ bodom joined02:14
← makopo left02:15
→ jar_corefile joined02:17
→ resmo__ joined02:17
← oc80z left02:19
← baburdick left02:19
← Error404NotFound left02:19
→ baburdick joined02:20
← spaceonline left02:20
→ oc80z joined02:20
← erichynds left02:20
← resmo_ left02:21
→ shiba_yu_ joined02:23
← shiba_yu36 left02:23
← Yuuhi left02:24
→ makopo joined02:25
→ Yuuhi joined02:26
→ bartek joined02:26
← univate_ left02:28
→ Error404NotFound joined02:31
← tjholowaychuk left02:33
→ caseymcg joined02:34
← dreiss left02:34
→ n8o-mba joined02:34
← cannonball left02:34
← brendon left02:35
→ maletor joined02:38
→ univate_ joined02:39
← univate_ left02:40
← MUILTFN left02:42
← infogulch left02:42
→ jezuz` joined02:42
← jar_corefile left02:43
→ mcluskydodallas joined02:43
← wilmoore left02:43
← cyphase left02:45
→ kvanderw joined02:45
→ cyphase joined02:46
← jezuz left02:46
← pheaver left02:47
← cyphase left02:47
← kermit left02:48
← shiba_yu_ left02:48
→ pheaver joined02:48
→ cyphase joined02:48
← mcluskydodallas left02:49
→ jlk joined02:49
← baburdick left02:49
→ wilmoore joined02:49
← eletuchy left02:50
← glennpratt left02:50
→ baburdick joined02:50
→ berserkr joined02:51
← catsup left02:52
→ neurodrone joined02:52
→ murilobr joined02:53
→ catsup joined02:53
→ kermit joined02:53
→ haseeb joined02:53
← berserkr left02:54
→ soulcake joined02:57
→ MUILTFN joined02:58
← kadoban left02:59
sitaram Mpenz: if you uncomment GL_PERFLOGT in the rc file you will get an additional log file that contains time taken for the request; that's about all you can get03:01
Mpenz: anything else has to written. And stuff like "how much memory did it use" or "how much network" etc., I don't know how to do -- anyone who knows can help03:01
← flamingspinach left03:02
→ kenperkins joined03:02
← jlk left03:03
Mpenz great, thanks sitaram03:04
:)03:04
I wonder if most existing deployments try to keep track of that or not?03:04
→ kadoban joined03:04
→ flamingspinach joined03:08
← linusoleander left03:09
→ angelsl joined03:12
← kukks left03:13
→ jalama joined03:17
→ ivorybishop joined03:19
← baburdick left03:20
→ osxorgate joined03:20
→ baburdick joined03:20
→ simplechat joined03:20
← simplechat left03:20
→ simplechat joined03:20
← n8o-mba left03:21
→ shiba_yu36 joined03:22
← jalama left03:23
sitaram Mpenz: only one person (see commit message for da210f2) wanted it. I basically refused to turn gitolite into a performance monitoring system, so I did whatever I could without touching core03:23
→ ivanoats joined03:25
← chrisf__ left03:28
→ n8o-mba joined03:28
→ chrisf__ joined03:28
← caseymcg left03:29
← joshwa left03:29
→ nullvariable joined03:30
← wilmoore left03:31
← angelsl left03:33
← baburdick left03:33
→ mlt- joined03:33
→ baburdick joined03:34
mlt- How do I find sha when a particular file was deleted?03:34
← aef left03:34
← makopo left03:34
cmn mlt-: git log --diff-filter=D -- path/to/file03:35
← osxorgate left03:35
→ aef joined03:35
mlt- cmn: thanks! it worked03:36
→ angelsl joined03:36
← warthog9 left03:36
← chrisf__ left03:37
← otakutomo left03:37
→ otakutomo joined03:37
mlt- now... can I somehow see where parts of content were moved from that deleted file?03:37
← baburdick left03:38
← pheaver left03:39
mlt- I feel like it should be something like blame -S ... but I can't just look for deleted file... instead I'd like to see what was affected by its content03:39
← MUILTFN left03:39
cmn mlt-: -C and/or -C -C03:40
mlt- cmn: it cannot stat that deleted file03:40
I've checked out that commit where file was deleted03:40
cmn IIRC, that looks hard for moved data, but it only works if you're looking at the file where it was moved to03:40
but you can just use git show03:40
← shiba_yu36 left03:41
mlt- i have a suspicion that that file was just gone :/ bummer03:41
→ makopo joined03:42
cmn git show <sha> where <sha> is the commit where the file was deleted will show the diff, if the contents were copied earlier, I'm not sure how to doit03:43
mlt- cmn: Thanks for help! I'm confident it was gone at once... hm... oh well... I was just checking if there were traces nearby commit with deletion03:45
← zivester left03:46
cmn you can bring it back with git checkout <sha> -- path/to/file03:46
mlt- cmn: it is not my code... I just had a patch so it worked but newer Ubuntu just don't have that code in a package. So I have to keep back old version :( That works at least03:47
→ berserkr joined03:48
← berserkr left03:49
← bartek left03:49
→ zivester joined03:50
← dorisabayon left03:52
→ JakeyChan joined03:53
→ MUILTFN joined03:55
← makopo left03:55
→ werdan7 joined03:56
← JakeyChan left03:58
→ warthog9 joined03:59
Mpenz sounds good.04:02
sounds good sitaram04:02
→ wilmoore joined04:02
← ISF left04:03
← OOPMan left04:03
→ JakeyChan joined04:04
→ makopo joined04:05
← warthog9 left04:06
→ cytrinox_ joined04:08
← cazic left04:08
→ mikepack joined04:11
← cytrinox left04:11
cytrinox_ → cytrinox04:11
← kenperkins left04:11
→ chrisf__ joined04:13
→ JakeyChan_ joined04:13
→ kenperkins joined04:13
← mikepack left04:14
← JakeyChan left04:16
JakeyChan_ → JakeyChan04:16
← dfr|mac left04:17
→ dfr|mac joined04:17
← dermoth_ left04:18
→ dermoth_ joined04:19
← mlt- left04:21
← dexter_e left04:24
→ shiba_yu36 joined04:26
→ univate joined04:27
→ towski joined04:29
→ Liquid-Silence joined04:29
← pdtpatrick left04:30
← shiba_yu36 left04:30
→ gchristensen joined04:31
→ AAA_awright joined04:31
← AAA_awright_ left04:33
← MacGyverNL left04:35
→ shiba_yu36 joined04:36
← xiangfu left04:36
→ tobago joined04:39
→ warthog9 joined04:40
← jaymurtii left04:40
→ LouisJB joined04:41
→ MacGyverNL joined04:41
← MacGyverNL left04:41
→ MacGyverNL joined04:41
← AAA_awright left04:43
→ AAA_awright joined04:44
← makopo left04:46
← gchristensen left04:48
→ makopo joined04:50
← makopo left04:52
→ boombatower joined04:52
→ defendguin joined04:53
defendguin is it possible to commit something via github i'm on a machine without git and i need to add a new file to a project04:53
?04:53
SamB I don't know if you can add a *new* file...04:55
defendguin hmm04:55
SamB but you can edit existing files!04:56
so go ahead and try04:56
defendguin i'm looking around don't see an upload option04:57
← univate left04:58
SamB I would be looking for a "new file" action04:58
failing that, maybe you can report a bug/issue and attach the file to that ?04:58
defendguin hmmm05:00
yeah let me know if you see a new file option05:01
← niloc132 left05:08
← ivanoats left05:08
← tobago left05:10
sgronblo So how are you suposed to solve conflicts on submodules?05:10
← orafu left05:11
← defendguin left05:11
→ orafu joined05:11
SamB sgronblo: git update --merge, it looks like05:13
er.05:13
git submodule update --merge05:13
← dfr|mac left05:14
← zivester left05:14
← lemonchicken left05:15
→ churp joined05:17
← shiba_yu36 left05:21
← emptyflask left05:21
→ lemonchicken joined05:24
→ apok joined05:27
→ amigojapan joined05:27
← tnks left05:28
→ Fandekasp joined05:28
→ tnks joined05:29
→ makopo joined05:32
← haseeb left05:33
→ mklappst_ joined05:34
→ bindaasomatic joined05:34
→ haseeb joined05:36
← mklappstuhl left05:37
← n8o-mba left05:37
← haseeb left05:38
← amigojapan left05:42
→ shiba_yu36 joined05:42
→ caseymcg joined05:47
← cherro left05:48
← ki__ left05:49
← mattt_ left05:50
→ maleknet joined05:53
→ mattt_ joined05:54
→ dexter_e joined05:55
→ dc5ala joined05:55
← subbyyy left05:57
← mattt_ left06:02
→ univate joined06:02
→ aidenhong joined06:03
→ pdtpatrick joined06:04
← dc5ala left06:06
→ dSebastien joined06:09
← btanaka left06:09
← makopo left06:12
→ dc5ala joined06:13
← shiba_yu36 left06:14
→ corburn joined06:15
→ amigojapan joined06:16
→ ehsan joined06:19
→ weiyang joined06:19
← wilmoore left06:19
→ mattt_ joined06:20
← Targen left06:20
← Bass10 left06:20
← bodom left06:23
← churp left06:23
← maletor left06:24
→ shiba_yu36 joined06:24
← kadoban left06:24
← corburn left06:25
→ OOPMan joined06:26
→ kurtzhong joined06:26
← dc5ala left06:26
→ Dave^| joined06:27
→ dc5ala joined06:27
← kurtzhong left06:28
→ drizzd joined06:29
← dc5ala left06:30
← JEEB left06:33
← neurodrone left06:33
→ tewecske joined06:33
→ JEEB joined06:34
← shiba_yu36 left06:34
← drev1 left06:35
← Samoi left06:39
← caseymcg left06:39
→ shiba_yu36 joined06:40
← shiba_yu36 left06:41
← chrisf__ left06:42
← LouisJB left06:46
→ LouisJB joined06:48
← tureba left06:48
→ tureba joined06:48
→ Raging_Hog joined06:50
→ tobago joined06:50
← OOPMan left06:52
← tureba left06:54
→ dc5ala joined06:54
→ chrisf__ joined06:56
← weiyang left06:56
→ ThomasLocke joined06:57
← mattt_ left06:58
→ baburdick joined07:00
← Aaaarg left07:01
← arvind_khadri left07:04
→ Chib joined07:05
→ na3r joined07:06
→ caseymcg joined07:08
← caseymcg left07:08
→ akostajti joined07:09
→ tureba joined07:10
← madewokherd left07:10
→ dr_lepper joined07:13
← dr_lepper left07:13
→ dr_lepper joined07:13
→ shurane joined07:13
→ krz joined07:19
→ ecthiender joined07:20
→ poseid joined07:21
← Liquid-Silence left07:21
→ Liquid-Silence joined07:21
→ whitman joined07:22
→ dangerousdave joined07:23
← akostajti left07:23
→ dreiss joined07:23
→ pheaver joined07:25
→ rendar joined07:26
← sneakret left07:27
→ SRabbelier joined07:27
→ Suresht joined07:28
→ krzkrzkrz joined07:29
← krz left07:30
krzkrzkrz → krz07:30
← tobago left07:31
→ pratz joined07:35
← chrizz left07:35
← pheaver left07:37
→ pheaver joined07:38
← towski left07:39
← pratz left07:41
→ jharrys joined07:42
→ shiba_yu36 joined07:43
→ kumbayo joined07:43
← s0ber left07:44
→ emptyflask joined07:44
→ s0ber_ joined07:44
s0ber_ → s0ber07:44
→ pratz joined07:45
← pratz left07:45
← kenperkins left07:46
→ pratz joined07:47
← jharrys left07:47
← dankest left07:49
→ makopo joined07:51
← MUILTFN left07:54
← LongBeach left07:55
→ tatsuya_o joined07:55
→ jharrys joined07:55
→ elb0w_ joined07:56
← noaXess left07:58
← elb0w left07:59
← coldboot left08:00
← Suresht left08:01
→ MUILTFN joined08:04
→ Targen joined08:05
← dexter_e left08:06
→ dexter_e joined08:07
→ coldboot joined08:08
← gebi left08:08
← tatsuya_o left08:08
→ TrineLocke joined08:10
→ jtr joined08:10
→ tvw joined08:12
← jtr left08:12
← LouisJB left08:13
← boombatower left08:14
← ecthiender left08:14
→ d0k joined08:16
← dexter_e left08:17
→ Spockz joined08:17
→ caseymcg joined08:18
→ DarkAR joined08:21
← lut4rp left08:26
← tvw left08:26
→ Daniel0108 joined08:26
→ tvw joined08:26
→ toobluesc joined08:27
→ lut4rp joined08:27
← lut4rp left08:27
→ lut4rp joined08:27
← dreiss left08:29
→ linusoleander joined08:32
→ dexter_e joined08:32
← linusoleander left08:32
→ qknight joined08:32
→ haseeb joined08:33
→ jamiemill joined08:33
← ajt left08:33
→ eletuchy joined08:34
← JakeyChan left08:34
→ thiago joined08:35
→ User390 joined08:36
← User390 left08:36
→ MissYeh joined08:37
← jharrys left08:39
→ ixti joined08:40
← haseeb left08:45
→ canton7 joined08:45
→ Cromulent joined08:46
→ pantsman joined08:47
← Cromulent left08:49
→ wks joined08:53
← dexter_e left08:53
← makopo left08:54
→ orsetto joined08:55
← canton7 left08:56
→ nimrod10 joined08:58
→ linusoleander joined09:00
← Raging_Hog left09:00
→ ecthiender joined09:00
← nimrod10` left09:01
← pantsman left09:01
→ pantsman joined09:03
← pantsman left09:03
→ pantsman joined09:03
→ haseeb joined09:03
← dhruvasagar left09:05
→ dhruvasagar joined09:05
← shiba_yu36 left09:07
← tvw left09:09
→ shruggar joined09:10
→ superdmp joined09:11
← pheaver left09:12
→ pdelgallego joined09:14
← dr_lepper left09:15
← aidenhong left09:15
← orsetto left09:16
→ LouisJB joined09:16
→ LongBeach joined09:17
← Fandekasp left09:17
→ Cromulent joined09:24
← _Vi left09:26
→ poga_ joined09:26
← Cogito left09:26
→ Cogito joined09:26
← poga_ left09:26
→ poga joined09:26
→ shiba_yu36 joined09:29
← shiba_yu36 left09:31
← caseymcg left09:32
→ shiba_yu36 joined09:33
← dSebastien left09:33
→ vdv joined09:33
← d4rk1ink left09:33
← shiba_yu36 left09:34
← Error404NotFound left09:34
← ecthiender left09:37
→ ecthiender joined09:37
→ sp3001 joined09:40
← kumbayo left09:41
← ecthiender left09:43
→ shiba_yu36 joined09:43
← techzin left09:43
→ techzin joined09:44
→ dSebastien joined09:45
← haseeb left09:46
→ afief joined09:47
→ dawkirst joined09:47
← flamingspinach left09:47
→ brian_g joined09:50
← bauruine left09:50
→ thomaschaaf joined09:55
← thomaschaaf left09:56
← Cromulent left09:58
→ ecthiender joined10:00
← canti left10:05
→ canti joined10:06
← shiba_yu36 left10:06
← pantsman left10:06
→ bgy joined10:06
bgy hi10:06
→ d4rk1ink joined10:07
bgy Is it possible with git-diff to show only the added part?10:07
→ psoo joined10:09
→ ecthiender_ joined10:10
← poseid left10:10
→ khmarbaise joined10:11
→ jimmy1980 joined10:11
← ecthiender left10:11
drizzd bgy: git diff|grep -v ^-10:12
bgy drizzd: clever :) I was looking for git option10:12
drizzd why? what difference does it make?10:13
→ pantsman joined10:13
← pantsman left10:13
→ pantsman joined10:13
bgy drizzd: no one, I said that was clever, I should have though about it but I was focused on a git option10:13
drizzd fwiw, maybe git blame output is more appropriate for what you want10:15
← dawkirst left10:15
drizzd it will have zeroes at the beginning of the line10:15
(or the corresponding sha1 if you are comparing to a commit)10:16
bgy drizzd: well it's gonna be used for a pre-commit hook, and I liked to only work on the added parts10:17
drizzd scratch that last part, it doesn't work for diffing commits10:17
→ martinjlowm joined10:18
← khmarbaise left10:18
→ sneakret joined10:19
→ _iron joined10:19
→ Chillance joined10:19
→ Samoi joined10:21
→ thomaschaaf joined10:21
← soulcake left10:26
← apok left10:28
→ heidar joined10:30
→ Hagbardceline joined10:30
→ dexter_e joined10:33
← jamiemill left10:34
→ mklappstuhl joined10:34
← likewhoa left10:34
→ apok joined10:35
← lut4rp left10:36
← Hagbardceline left10:37
← cypha left10:38
← RevoOf left10:39
← dhruvasagar left10:42
→ dhruvasagar joined10:42
← univate left10:44
← Spockz left10:46
thomaschaaf Hey I am very new to git. I keep getting: http://pastie.org/2699848 when I do git submodule update :( can someone help my understand what the problem is? I have a gitmodules file which should download https://github.com/twitter/bootstrap10:47
into vendor/twitter-bootstrap10:47
→ subbyyy joined10:50
thomaschaaf doing git status in vendor/twitter-bootstrap gives me a bunch of deleted files int that folder :(10:51
but why?10:51
← apok left10:53
← dexter_e left10:55
→ dexter_e joined10:56
→ Textmode joined10:58
→ knifed joined10:58
→ steffo joined11:03
→ aidenhong joined11:04
→ apok joined11:07
→ JakeyChan joined11:08
← DarkAR left11:08
→ kerx joined11:09
← chrisf__ left11:10
kerx hi all, i'm getting: fatal: Unable to find remote helper for 'https', when I try to do a 'git clone'. I compiled git with the following configure: ./configure --prefix=/opt/git --with-openssl --with-curl --with-expat && make11:10
any ideas, as to why this would be happening?11:10
thiago try http11:10
kerx same11:11
fatal: Unable to find remote helper for 'http'11:11
thiago look at your installation's libexec/git-core. Which executables with "http" in the name do you have there?11:12
← amigojapan left11:14
kerx thiago, I'm re-building again, I will give you that info in ~2 minutes11:15
← dhruvasagar left11:16
kerx thiago, the only thing I have in there with 'http' is git-http-backend11:16
← superdmp left11:16
→ dhruvasagar joined11:16
thiago that's not the one you need11:17
kerx what could be causing my installation not to install the proper binaries for http and https?11:17
thiago I'm not sure if it's git-http-fetch or git-remote-http that you need11:17
those binaries not being built at all11:17
kerx I have git-fetch11:17
← whitman left11:18
kerx I have both --with-curl, and --with-expat turned on11:18
← steffo left11:18
→ Raging_Hog joined11:20
→ beautifulmind joined11:23
← pantsman left11:23
← MUILTFN left11:23
→ _md joined11:28
→ jakobm joined11:29
← dhruvasagar left11:29
→ dhruvasagar_ joined11:30
→ ssb joined11:31
← murilobr left11:31
ssb hi! how do I figure out, which local and remote branches have a particular commit?11:32
thiago git branch -a --contains <commit here>11:32
ssb thiago, *blush* thanks! dunno how did I miss that in man git-branch11:33
jast thiago: *automatic message* the 'git-branch' manpage can be found at http://www.kernel.org/pub/software/scm/git/docs/git-branch.html11:33
← otakutomo left11:33
→ otakutomo joined11:34
→ jamiemill joined11:36
→ munichlinux joined11:38
diverdude When i do git commit it seems that git does not detect changes in files. It just says: no changes added to commit even though i have changed the content of some files. Why is that?11:38
← LouisJB left11:38
cbreak diverdude: because git doesn't commit your changes11:39
thiago diverdude: design choice11:39
cbreak it commits your staged changes11:39
read man git-add and man git-commit11:39
jast *automatic message* the 'git-add' manpage can be found at http://www.kernel.org/pub/software/scm/git/docs/git-add.html11:39
→ berserkr joined11:39
diverdude cbreak, ohh ok...but how does it make sense to not track changes? Isnt that the whole idea of a versioning system?11:40
→ j416 joined11:40
cbreak no11:40
diverdude i have file...its version 1...as soon as i type something in the file, its a new version...How can that not be a central part of a versioning system11:41
thiago diverdude: design choice11:41
diverdude: without arguments, git commit commits only what you git added11:41
diverdude thiago, but i cannot sit and tell git everything i change in my files11:42
thiago diverdude: you can11:42
diverdude how11:42
thiago git add the file, or give git commit the file name11:42
diverdude thiago, i did those11:42
cbreak diverdude: are you kind of slow?11:42
thiago no, you didn't11:42
cbreak read man git-add and man git-commit11:42
jast *automatic message* the 'git-add' manpage can be found at http://www.kernel.org/pub/software/scm/git/docs/git-add.html11:42
diverdude thiago, i did git add .11:42
thiago diverdude: do it again11:43
diverdude: before every commit11:43
diverdude really?11:43
thiago yes11:43
cbreak careful with git add .11:43
it doesn't do the right thing in many situations11:43
diverdude i have to add . b4 every commit?11:43
cbreak no11:43
thiago no11:43
→ akostajti joined11:43
thiago you have to git add the files you want to commit11:43
cbreak use your brain11:43
thiago or you can tell git commit which files to commit11:43
cbreak stage the things you want to commit11:43
not everything in the current folder11:44
thiago or you can use one of the options to git add or git commit that tell them to search for modified files11:44
diverdude but if i change 20 files between each commit do i have to do git add file1;git add2 file2 etc?11:44
← fisted left11:44
thiago diverdude: read what I've just said11:44
cbreak there are tools like git add -p, git add -u, git add -A and git commit -a that help11:44
diverdude: don't play dumb11:44
I know you know what to do11:44
it's right in the man page!11:44
diverdude sounds quite insane...maybe i just dont get the idea11:45
thiago it was a design choice11:45
and it allows git add -p to exist11:45
cbreak I think it was a very good choice11:45
diverdude sounds like a really strange choice11:45
cbreak it allows changing two things and committing them separately11:45
unlike that pile of garbage that is svn11:45
where you have to somehow diff patch stuff away11:45
← psoo left11:45
cbreak and hope it still applies after committing11:46
diverdude in svn i just du svn commit and thats it11:46
cbreak what a pain in the ass11:46
thiago diverdude: suppose you modified two things in one file11:46
diverdude: how do you commit just one of the two things?11:46
← berserkr left11:46
diverdude thiago, i dont want to...i want to commit both things since they are both part of the new version...otherwise i would have committed earlier11:47
cbreak no11:47
thiago diverdude: let me change the discussion:11:47
cbreak think a tiny bit11:47
thiago diverdude: suppose you *want* to commit one of the two things only11:47
diverdude thiago, when wold you ever want that?11:47
cbreak you make some experimental changes and you find a bug11:47
you fix the bug11:47
diverdude *would11:47
thiago because people don't think linearly11:47
cbreak you don't want to commit the experimental changes of course11:47
thiago when I'm working on A, I sometimes do B, C, and D.11:47
cbreak because they aren't yet ripe nor tested11:47
thiago when I'm done doing what I was working on, I realise that the change is composed of smaller parts11:48
→ fisted joined11:48
thiago a concrete example: refactor some code changing the API11:48
diverdude so its possible to commit only parts of a file?11:48
cbreak yes11:48
of course11:48
it'd suck if you couldn't11:48
thiago when you're done, the old code ceased to exist, the new code is now in existence and there are changes throughout the codebase changing to the new API11:48
diverdude do you do commit myfile line 4-40 or how do you specify that?11:49
cbreak git add -p -- filename11:49
thiago those are three separate commits: 1) add the new code; 2) change the rest of the codebase to use the new API; 3) delete the old code11:49
diverdude git add -p -- filename ... how does it know what part of the file to commit then11:50
cbreak are you stupid?11:50
diverdude yes i am'11:50
cbreak what do you think git does when you answer "yes" or "no"?11:50
thiago diverdude: because you tell it11:50
diverdude: git add -p will ask you what to do. It's interactive.11:51
diverdude hmm ok11:51
sounds like commit has terribly cumbersome in git11:51
cbreak haha...11:51
thiago no, it's pretty easy11:51
cbreak it's pretty easy now :)11:51
→ johnkpaul joined11:52
cbreak just think back to what pain you had to suffer to commit something in svn11:52
some projects I work on still use that... it's terrible :(11:52
thiago when you want to commit, git add the file(s) you want to commit11:52
diverdude cbreak, i seem to remember that svn commit was pretty easy11:52
thiago use -p if you want to add only a portion of the file (partial add)11:52
and please pay attention to this:11:52
cbreak diverdude: because you never really committed anything I guess11:52
→ DarkAR joined11:52
thiago if you want, you can tell git add or git commit to scan for all modified files11:53
diverdude cbreak, ehhh ok right11:53
thiago cbreak: stop insulting him11:53
cbreak I had to fix a heap of conflicts last week to even be able to commit11:53
← bolt left11:53
cbreak and first I had to remove experimental changes from the working directory11:53
and after the commit I had to put them back in11:53
took me over 10 mins11:54
with git I do such things in a few seconds11:54
← johnkpaul left11:54
diverdude thiago, ok, that sounds like something which would be useful11:54
cbreak even git-svn makes this job easier11:54
diverdude hmm ok11:55
cbreak (because you can commit locally first, without having to wrestle with the svn server)11:55
diverdude: if you want svn style all-committing, do as I told you three times or so, and read man git-commit to find git commit -a11:57
jast diverdude: *automatic message* the 'git-commit' manpage can be found at http://www.kernel.org/pub/software/scm/git/docs/git-commit.html11:57
Kicer86 hi all11:57
i've a git repos on two machines, one is with windows, one with linux.11:57
I'd like ot create repo on pendrive to move data between those two computers11:57
→ jaymurtii joined11:57
Kicer86 one problem: pendrive must be encrypted11:57
cbreak truecrypt? :)11:58
→ arvind_khadri joined11:58
Kicer86 and i don't want to instal truecrypt11:58
;)11:58
cbreak there's a portable version11:58
needs admin rights on windows though11:58
(not sure if there's anything similar on linux)11:58
thiago Kicer86: there are two steps:11:58
Kicer86: 1) find an encryption mechanism that works on Windows and Linux11:59
Kicer86: 2) create a Git repository on it11:59
→ the_metalgamer joined11:59
thiago Kicer86: we'll help you with step 2 only.11:59
Kicer86 the problem is that i don't like when dome 3rd party programs are messing in my linux machine... ;)11:59
heh ;)11:59
thiago Kicer86: let us know when you have found the encryption mechanism.11:59
cbreak git is a third party program11:59
Kicer86 but it doesn't mess deep in my linux11:59
cbreak heck, even gnu are third party programs11:59
Kicer86 it is just installed via package12:00
that's ok12:00
cbreak why is truecrypt not ok? :)12:00
Kicer86 it's not installed via package ;P12:00
cbreak uhm... make a package for it12:00
then it's installed via pckage12:00
easy12:00
Kicer86 ;)12:00
i'm looking for ready solutions ;)12:00
cbreak truecrypt12:01
thiago Kicer86: this is the wrong channel to ask...12:01
cbreak the only cross platform FS encryption I know of12:01
there are probably file encryption tools, but those are obviously unsuitable for git12:01
Kicer86 thiago: i was hoping to hear: new git has implemented encryption ;)12:01
cbreak unless you encrypt each git-bundle separately12:01
thiago Kicer86: unfortunately, it hasn't happened12:01
cbreak Kicer86: git doesn't care about things like authentication or security12:02
Kicer86 cbreak: indeed12:02
too bad12:02
ok, truecrypt then...12:02
cbreak you could switch to OS X12:02
and scrap windows and linux12:02
then you could use built in encrypted disk images :D12:02
no installing neccessary!12:02
thiago linux has built-in encrypted images too12:02
cbreak (or windows with bit locker)12:02
Kicer86 ;)12:02
cbreak it has?12:02
thiago sure12:02
LUKS12:02
cbreak I thought in linux they are third party12:02
Kicer86 cryptoloop ;P12:03
thiago let me see... when did I buy that 500 GB disk...12:03
cbreak if it's not hand crafted by linus himself then I don't want it!12:03
thiago since at least 200812:03
all my external hard drives are encrypted now12:03
← flavius left12:03
thiago my new laptop has an encrypted root and encrypted swap12:03
the only thing that's not encrypted is /boot12:04
cbreak I use the OS X encryption. It's easy to use and probably more secure than nothing12:04
thiago uses LUKS12:04
thiago I measured almost no performance loss due to it12:04
cbreak I doubt there's performance loss here considering that most of my cores are idle most of the time...12:05
thiago a Qt build with encrypted fs took 27 minutes, while one without encryption took 2612:05
→ Adys joined12:05
cbreak compiling is probably IO bound12:05
but encryption doesn't add much IO overhead12:05
have you tried to compile Qt on an SSD? :)12:06
← fisted left12:06
thiago this is an SSD12:06
→ whitman joined12:06
→ krisoijn joined12:06
← the_metalgamer left12:07
← Dreamer3 left12:07
Kicer86 thiago: how long do you have it?12:08
(SSD)12:08
← ecthiender_ left12:09
← Daniel0108 left12:09
→ ecthiender joined12:09
→ guy_brush joined12:10
→ univate joined12:13
← univate left12:13
→ sabgenton joined12:15
← krisoijn left12:15
thiago Kicer86: 3 months12:15
sabgenton is there any chance git will move more to c based code to the point the bash and perl stuff will almost go away?12:15
Kicer86 oh, i'm looking for some longer owners12:15
thiago sabgenton: moving to more C based code has already happened12:16
Kicer86 to make sure how ssd disks behave then they get older12:16
sabgenton yes but will it go futher?12:16
thiago sabgenton: for some programs, it might12:17
→ gusnan joined12:17
sabgenton I just fine it a bit anyoning on windows12:17
→ Fandekasp joined12:18
→ fisted joined12:19
sabgenton tortoise svn seemed to build there own svn core I just wonder if that will happen with git on various platforms?12:19
thiago why would anything happen?12:20
→ the_metalgamer joined12:21
sabgenton thiago: to have a none commandline environment option12:21
← munichlinux left12:22
→ phreaknerd joined12:25
← beautifulmind left12:29
→ xiangfu joined12:29
→ poseid joined12:31
→ bartek joined12:32
→ akosikeno joined12:32
← ehsan left12:32
← akosikeno left12:33
→ akosikeno joined12:34
← jakobm left12:35
← akosikeno left12:35
→ akosikeno joined12:36
← akosikeno left12:36
← ecthiender left12:37
→ jakobm joined12:37
→ akosikeno joined12:37
← dexter_e left12:37
← ixti left12:38
← akosikeno left12:38
→ akosikeno joined12:39
→ neurodrone joined12:40
← akosikeno left12:40
→ akosikeno joined12:41
sitaram sabgenton: jgit is git in java12:41
and there are implementations in other languages as well, I think12:41
← akosikeno left12:42
sitaram I dont know how good/bad they are though. And I personally have no sympathy for "non-command line" requests, so I probably may not have kept track of such12:42
sabgenton hmm yeah I remmber that now I have a friend using eclipse who tried it12:42
→ akosikeno joined12:42
sitaram if you're a developer, and you're afraid of the command line, you're no use to me... <-- just my personal opinion; not to be taken as a generalisation12:43
sabgenton thats a good example though12:43
sitaram (and "you" not to be taken as you personally!!!)12:43
← akosikeno left12:43
sabgenton sitaram: I like the commandline12:43
just nobody else does that I associate with12:44
I personally don't think it is everything12:44
← shruggar left12:44
→ akosikeno joined12:44
← arunce left12:45
← otakutomo left12:45
← akosikeno left12:45
sitaram sabgenton:\\nathanj is a very good guide to using git entirely through the GUIs that come with it12:45
oope12:45
sabgenton like I think git as a distrusted file system / vcs is fairly revolutionary12:45
→ otakutomo joined12:45
sitaram sabgenton: http://nathanj.github.com/gitguide/ is a very good guide to using git entirely through the GUIs that come with it12:45
sabgenton but the platform just 'happens' to be on the commandline as far as I care12:46
sitaram "distrusted"?12:46
→ akosikeno joined12:46
m1sc there's dulwich, written in python, apparently used by google code12:46
sitaram anyway you should point your friends to that link I just posted12:46
m1sc sabgenton: you are aware of libgit2?12:47
← akosikeno left12:47
→ akosikeno joined12:48
sabgenton no interesting thanks12:49
← akosikeno left12:49
sabgenton sitaram: I'm swaring myself to cygwin and commandline only12:49
untill there is something better than msysgit12:49
→ akosikeno joined12:50
← akosikeno left12:50
sitaram sabgenton: my Windows users only use msysgit; I prefer it to cygwin for install ease. Anyway we dont run servers on Windows so the only additional advantage of cygwin is nullified12:50
← krz left12:51
→ akosikeno joined12:51
→ Daniel0108 joined12:52
sabgenton I just like being able to push pull on cygwin like I was on a linux station12:52
← akosikeno left12:52
← kerx left12:53
← apok left12:53
sabgenton aparently linux projects with permisons don't get munched in cygwin git12:53
→ akosikeno joined12:53
← thomaschaaf left12:53
sitaram has no idea what that means12:53
← ssb left12:53
sabgenton file modes aren't preserved in msysgit12:54
→ arunce joined12:54
sabgenton apparently12:54
not that I've got into it12:54
← pdelgallego left12:54
← arvind_khadri left12:55
← jakobm left12:55
m1sc sabgenton: did you discuss this with the msysgit people? i can't image; wouldn't allow any serious cross platform dev teams.12:56
imagine12:57
sabgenton http://www.grumpydev.com/2011/01/19/switching-from-cygwin-to-msysgit-git-thinks-everything-has-been-modified/12:58
→ Vortex35 joined12:58
sabgenton m1sc: anyrate I hop libgit2 becomes part of a working product that would be awesome12:58
→ Bass10 joined12:59
← topping left13:01
← guy_brush left13:02
← Vortex34 left13:02
m1sc sabgenton: afaik TortoiseGit is using it for example.13:03
← MissYeh left13:03
sabgenton no13:03
it use msysgit13:03
it's just a fronted I think13:03
m1sc sabgenton: i should have said in parts - libgit2 is still lacking some core functions..13:06
→ munichlinux joined13:06
sabgenton well as long as it never dies I'm happy :)13:06
thx for the tipoff13:07
→ arvind_khadri joined13:07
→ MissYeh joined13:08
← Daniel0108 left13:08
← xiangfu left13:10
sitaram sabgenton: I looked at that link you posted about msysgit etc. I have to tell -- mixed environment projects will always have such problems, and the only reason cygwin appears to succeed is that it apparently provides a more complete Unix emulation13:10
→ whit537 joined13:10
sitaram sabgenton: my question is: why emulate? Why not use the real thing?13:10
anyway it's a somewhat rhetorical question; feel free to ignore13:11
→ CannedCorn joined13:11
→ ehsan joined13:12
← c00kiemon5ter left13:13
→ tatsuya_o joined13:14
→ mattt_ joined13:14
sabgenton sitaram: well yeah if I was doing a total windows project then yes why would I. The cygwin shell was more in hopes that I could not break things I made on *nix machines13:15
→ c00kiemon5ter joined13:15
← c00kiemon5ter left13:15
→ c00kiemon5ter joined13:15
sitaram thinks mixed env projects should be *developed* in Unix and merely *tested* in Windows. If the developers are using Windows... <sigh>13:16
→ glennpratt joined13:16
→ bolt joined13:19
← arvind_khadri left13:19
wereHamster agrees13:20
→ nicoulaj joined13:20
→ chrizz joined13:22
→ amcsi_work joined13:23
amcsi_work hello13:23
← eletuchy left13:23
amcsi_work is there a shorter version of git push . HEAD:<branchname>?13:23
← CannedCorn left13:24
→ univate joined13:25
wereHamster why would you do that?13:26
→ OrionX joined13:26
← univate left13:26
wereHamster and also, how short do you want it to be?13:26
← poseid left13:26
→ Orion_X joined13:26
→ ejnahc joined13:28
sitaram what does that command even *do*... sounds too arcane to be sane13:29
← afief left13:29
→ afief joined13:30
← OrionX left13:30
wereHamster sitaram: git checkout branchname && git merge HEAD@{1}. kindof, except it only updates the branch if it's ff13:31
→ arvind_khadri joined13:31
← jimmy1980 left13:31
wereHamster one 'shortcut' would be to use update-ref, which doesn't do the ff check.13:31
← localhost left13:31
← sabgenton left13:31
sitaram wereHamster: oh yeah; I parsed it correctly now; thanks!13:32
wereHamster: and if you are willing to not do the ff check, 'git branch -f branchname HEAD' is more porcelain than update-ref13:33
I imagine that would do the same13:33
amcsi_work: in general, whenever anyone asks for a "shorter" version, it means they havent thought about gitconfig aliases, shell functions, etc13:34
→ localhost joined13:34
amcsi_work okay, I rather mean a more logical version13:35
like for instance13:35
git reset <commit-like> makes the current branch point to commit-like13:36
→ kukks joined13:36
wereHamster that's what branch -f does13:36
(minus the non-ff check)13:36
amcsi_work but when I'm not on the branch that I want to point to somewhere else, I get a feeling there should be a command specifically to repoint a branch to another commit, without having to reference dot13:37
→ jimmy1980 joined13:37
wereHamster amcsi_work: that's what branch -f does13:37
amcsi_work yes, but branch -f can only move a branch to the current commit13:37
wereHamster no13:37
amcsi_work wait, actually, you're right13:37
wereHamster git brnach -f whateverbranchyouwanttopoint tothiscommithere13:37
→ Jester01 joined13:37
amcsi_work that is equivalent to what I asked13:38
wereHamster no13:38
brnach -f doesn't perform the non-ff check13:38
amcsi_work git push . HEAD:<branchname>13:38
wereHamster that will not update <branchname> if it's not a ff update13:38
amcsi_work oh I see13:38
← ejnahc left13:39
→ madewokherd joined13:40
sitaram he would have '-f'd it if he'd know :)13:40
Jester01 hi people! I have done a git rebase where I skipped a patch. turns out, I'd still need to apply that. what's the easiest way to get it back?13:40
wereHamster Jester01: git reflog13:40
→ ejnahc joined13:42
← techzin left13:42
Jester01 thank you13:43
← TrineLocke left13:45
← RaceCondition left13:45
← drizzd left13:46
← wagle left13:46
→ alansaul joined13:52
→ Daniel0108 joined13:52
← aidenhong left13:55
→ aidenhong joined13:56
← tatsuya_o left13:56
← cenuij left13:59
→ the_cybersphinx joined14:02
→ balau joined14:03
← afief left14:04
→ afief joined14:04
→ ginabc joined14:04
← cybersphinx left14:06
→ murilobr joined14:06
← ginabc left14:09
→ mandric joined14:09
← ThomasLocke left14:09
→ gbacon joined14:12
→ Error404NotFound joined14:13
← Error404NotFound left14:13
→ Error404NotFound joined14:13
← gbacon left14:13
← Chib left14:13
→ zivester joined14:13
← maleknet left14:13
→ maleknet joined14:13
← tchan left14:14
→ oriba joined14:15
← pratz left14:16
→ pratz joined14:17
← pratz left14:17
→ dfr|mac joined14:18
→ giu joined14:18
← maleknet left14:19
← afief left14:20
→ afief joined14:20
→ davzie joined14:20
→ wagle joined14:20
← mandric left14:21
← davzie left14:22
← jimmy1980 left14:22
→ macUzer joined14:22
← alansaul left14:23
← giu left14:24
→ KanaY joined14:25
→ jakobm joined14:25
← bartek left14:26
← MissYeh left14:26
← Daniel0108 left14:28
→ jimmy1980 joined14:28
← akostajti left14:29
→ jalama joined14:30
← munichlinux left14:30
→ Daniel0108 joined14:31
→ dexter_e joined14:32
← JakeyChan left14:33
← wks left14:33
← na3r left14:34
sitaram Mpenz: http://dpaste.org/MhzXT/ <-- some performance metrics from my 3+ year old laptop14:35
away ...14:35
→ joneskoo joined14:35
← phreaknerd left14:35
← dexter_e left14:36
← kvanderw left14:36
← knifed left14:39
→ JakeyChan joined14:40
← arvind_khadri left14:44
→ ReekenX joined14:46
← Drakonite left14:47
→ cakehero joined14:47
← jalama left14:48
← ptl left14:50
→ muneeb joined14:51
← jaymurtii left14:51
← otakutomo left14:51
→ otakutomo joined14:51
← muneeb left14:52
→ nixmaniack joined14:52
→ Drakonite joined14:54
→ kvanderw joined14:54
← Textmode left14:55
→ Guest46173 joined14:55
← Guest46173 left14:56
→ arvind_khadri joined14:57
→ sikao_lfs joined14:57
→ tatsuya_o joined14:58
→ swombat joined14:59
→ dexter_e joined14:59
jefferai Does anyone have any good reference for the difference between the Git and Hg branching models? http://whyisgitbetterthanx.com says that the branching model in Git is unique among all of them, but the model it describes is, I thought, basically the same way that Hg does it15:01
→ niloc132 joined15:01
jakobm jefferai: search for the article by steve losh on branching with mercurial15:01
jefferai jakobm: awesome, thanks15:02
→ erichynds joined15:02
← arvind_khadri left15:04
→ poisonbit joined15:04
→ lenswipe joined15:06
← lenswipe left15:06
→ lenswipe joined15:06
lenswipe hey guys15:06
how do i push to a remote github repo15:06
← JakeyChan left15:06
← arunce left15:06
lenswipe if i do git push origin i get: You can't push to git://github.com/lenswipe/CodeIgniter.git15:06
Use [email@hidden.address]15:06
Marquel lenswipe: you have to set up your remote url correctly for that.15:07
lenswipe: see the output of git remote -v15:07
lenswipe so how do i push back to github?15:07
→ JakeyChan joined15:07
jefferai jakobm: there's some decent misinformation on there, but it doesn't really answer my question, unfortunately15:07
in the whyisgitbetterthanx site it says "Probably Git's most compelling feature that really makes it stand apart from nearly every other SCM out there is its branching model. It is completely different from all of the models I'm comparing it to here, most of which recommend that the best branch is basically a clone of the repository in a new directory."15:07
Marquel lenswipe: git remote can also set the push-url to the required type.15:08
lenswipe i forked a repo, downloaded it and made some changes now i want to commit it to the remote repo15:08
jefferai and it labels that as a plus over Hg and Bzr15:08
but I don't really see a big difference15:08
lenswipe when i do a commit it just commits to my local repo15:08
how do i fix this?15:08
← afief left15:08
→ afief joined15:08
Marquel lenswipe: not at all.15:08
lenswipe: commits are always local only. pushing is the right way to go. but as i've said: the remote url needs to be set correctly and it seems it isn't.15:09
jakobm jefferai: don't trust whyisgitbetterthanx, do you have a specific question regarding the branching differences?15:09
→ kenperkins joined15:09
jefferai nope, I was just trying to understand that statement15:09
lenswipe Marquel, so how do i set it?15:09
Marquel see above: git remote is your friend.15:10
→ towski joined15:10
→ w0bni joined15:13
lenswipe robert@C03:~/CodeIgniter$ git push lenswipe/CodeIgniter15:13
fatal: 'lenswipe/CodeIgniter' does not appear to be a git repository15:13
fatal: The remote end hung up unexpectedly15:13
→ Ff2 joined15:13
← Ff2 left15:13
seed419 hey how do I commit a deletion to also delete a file on the repository15:14
Marquel lenswipe: did you set up the push-url to exactly what the error said? ("git@github.com:/lenswipe/CodeIgniter.git")15:15
lenswipe i fixed it15:15
yeah :)15:15
cmn seed419: if you want to remove the file completely from all the history, you need to use filter-branch, but this is against what version control is about15:15
lenswipe Marquel, how do i delete the "nbproject" file from github?15:16
seed419 its an old file that i don't want people accidentally downloading15:16
→ holybit joined15:16
← Error404NotFound left15:16
Marquel lenswipe: no idea, i am not using github or anyhow related to git at all. i'm just answering questions i've had myself some time ago.15:17
lenswipe O_O15:17
ok15:17
how would i delete a file from a remote repo then?15:17
seed419 that's what I'm asking :)15:17
← ReekenX left15:17
lenswipe lol15:17
→ arvind_khadri joined15:17
bgy how could I use multiple pre-commit hooks in git? Do I need to deploy several hook and call them all in .git/hooks/pre-commit?15:18
→ brendon joined15:18
_ikke_ bgy: Yes15:19
cmn seed419: you'd have to rewrite pretty much all of the history from the moment that file entered the repo15:19
→ mandric joined15:19
← kenperkins left15:19
Marquel lenswipe: git rm will remove the file, then commit and push your change.15:21
lenswipe ok15:21
→ ReekenX joined15:21
Marquel everything else is against the target of version control.15:21
lenswipe i dont want to delete it here, i just want to remove it from the local repo15:21
its a file of metadata from netbeans15:21
its not private sutff, its just un-neccecary15:21
Marquel is it committed?15:21
lenswipe yep15:22
Marquel then you need to git rm it anyway.15:22
← Jester01 left15:22
lenswipe ok15:22
will that delete it locally?15:22
Marquel if you don't want it to reappear: put a filemask into .git/info/excludes15:22
yes. so better keep a copy if you still need it.15:22
→ haseeb joined15:22
lenswipe ok15:23
seed419 yeah that works, thanks :)15:23
← sikao_lfs left15:25
Marquel .git/info/excludes is not pushed or something like that, so you have to keep it, if you recreate your local clone(s).15:25
→ chaoflow joined15:27
→ topping joined15:28
← arvind_khadri left15:28
→ RaceCondition joined15:28
→ kenperkins joined15:28
← amcsi_work left15:30
← RaceCondition left15:32
← erichynds left15:32
lenswipe Marquel, just did git rm, the file is still there in the remote repo15:34
→ shruggar joined15:35
Marquel lenswipe: did you commit and push?15:37
lenswipe yes15:37
Marquel lenswipe: it will not disappear from the history.15:37
lenswipe like i said15:37
its not private15:37
just irrelevant15:37
there its gone now :D15:38
Marquel, so if i want to keep it here without it being copied back into the repo do i just add a .gitignore file?15:39
← haseeb left15:39
Marquel .git/info/excludes for each local clone where it is and should not be committed.15:40
probably .gitignore globally, dunno about that.15:40
lenswipe ok so how do i add a file to .git/info/excludes?15:40
→ RaceCondition joined15:40
Marquel open .git/info/excludes and add a line - relative path, wildcards allowed like in any shell command.15:41
so if it is a file in top-level, the filename is perfectly sufficient.15:41
SamB Marquel: except it's not like in shell commands, since some are taken to apply everywhere in the tree and some aren't...15:42
lenswipe and nbproject would exclude the whole nbproject folder?15:42
recursively15:42
Marquel nbproject/ it should be ;)15:42
lenswipe ah15:42
→ munichlinux joined15:42
← RaceCondition left15:42
← kvanderw left15:45
← kenperkins left15:46
→ sorin joined15:46
← ejnahc left15:48
→ w0bni1 joined15:48
→ ejnahc joined15:49
→ RaceCondition joined15:49
← khaije|minder left15:50
← w0bni left15:50
→ cypha joined15:50
→ haseeb joined15:52
→ kadoban joined15:53
← ejnahc left15:53
← ReekenX left15:53
→ ejnahc joined15:53
← gretch left15:53
→ gretch joined15:54
← RaceCondition left15:54
→ jalama joined15:54
→ arunce joined15:55
chaoflow git blame -C -C -C does not work: https://gist.github.com/1289745 - am I or is git guilty?15:55
← JakeyChan left15:57
→ amigojapan joined15:58
← cakehero left15:58
cmn chaoflow: try with about 10 lines or so, what you're giving it is way too little context15:59
← EvilDaniel0108 left15:59
→ jondot joined16:00
← jbalint_ left16:00
← Cogito left16:00
→ Cogito joined16:00
→ likewhoa joined16:01
ChanServ set mode: +v16:01
← tatsuya_o left16:01
jondot hi guys, i'm looking for a good GUI for git under ubuntu. the purpose is to introduce microsoft-stack TFS devs to git, and _not_ dropping them to cmdline first, any tips?16:01
Marquel on windows i'd say tortoisegit, but linux...16:02
→ Prasad joined16:02
← topping left16:03
← vdv left16:04
jondot yep, that's my question16:06
← jalama left16:07
Marquel well... you might try gitg, which is a gnome-app.16:07
cmn there's gitk and git gui16:07
→ kenperkins joined16:07
Marquel but, tbh. i use it only to visualize the branches, not to do really use it for committing.16:07
→ tatsuya_o joined16:07
jondot hm. ok i guess an ide would make that integration better.16:09
→ EvilDaniel0108 joined16:09
Marquel well, i use kdevelop as IDE and git-integration into that is quite good.16:09
cmn EGIT is for Eclipse and works on anything16:10
→ soc42 joined16:11
chaoflow cmn: thx, that did it!16:11
→ khmarbaise joined16:12
Marquel was about suggesting git for eclipse too16:12
← towski left16:14
→ dreiss joined16:14
← jamiemill left16:14
→ Textmode joined16:15
m1sc jondot: what problem do you see with using straight cmdline?16:15
← dexter_e left16:15
jondot m1sc, target audience :)16:15
→ LouisJB joined16:17
→ cannonball joined16:17
← DarkAR left16:18
← nicoulaj left16:19
← sorin left16:21
← macUzer left16:22
← oriba left16:23
← ivorybishop left16:23
→ nicoulaj joined16:24
← linusoleander left16:25
→ RaceCondition joined16:26
→ sythe joined16:26
→ linusoleander joined16:26
→ Spockz joined16:29
→ towski joined16:29
→ bartek joined16:30
→ jalama joined16:31
→ Cromulent joined16:31
← jalama left16:31
← Cromulent left16:33
→ Zelest_ joined16:34
← munichlinux left16:34
→ alvaro_o joined16:35
← alvaro_o left16:35
← Zelest left16:37
→ ivorybishop joined16:37
→ ISF joined16:37
→ mikepack joined16:37
← bartek left16:37
← simplechat left16:37
→ rewire joined16:39
← towski left16:41
→ [1]soc42 joined16:41
← khmarbaise left16:42
→ _psychic_ joined16:43
← soc42 left16:44
→ Cromulent joined16:44
→ munichlinux joined16:45
← Zyrkon left16:46
← troyt left16:46
← mikepack left16:47
← Cromulent left16:49
← munichlinux left16:51
→ DarkAR joined16:51
← dhruvasagar_ left16:51
→ troyt joined16:52
← KanaY left16:52
obiwahn if tortoise-git is as crappy as tortoise-svn then i would not suggest using it:)16:52
→ sorin joined16:52
obiwahn anybody who considers himself a developer should put some time into learning something about scms16:54
→ dhruvasagar joined16:54
→ i42n joined16:54
→ soc42 joined16:55
→ drev1 joined16:55
→ munichlinux joined16:56
← i42n left16:56
→ Joghurt joined16:56
→ tchan joined16:57
← [1]soc42 left16:57
← Prasad left17:00
← nicoulaj left17:01
← haseeb left17:04
→ superdmp joined17:07
→ NaNoNickV joined17:07
→ dim_ joined17:07
→ harry23 joined17:08
cbreak tortoise-git is even more crappy than tortoise-svn17:08
tortoise-svn is legitimately full of SVNisms17:08
but in tortoise-git those are out of place17:08
← NaNoNickV left17:09
→ Zyrkon joined17:11
→ mikepack joined17:11
rewire git extensions http://code.google.com/p/gitextensions/ is prob the best git-gui there is :D17:12
← neurodrone left17:12
obiwahn i guess his attempt is futile - when they are used to msvc only there is not a chance:P only thing that can be done is offer people who are sick of msvc alternatives .. forcing people will not succeed.17:13
→ kadoban_ joined17:16
← kadoban left17:17
← whit537 left17:19
→ cytrinox_ joined17:20
← _psychic_ left17:20
← murilobr left17:21
jepler- → jepler17:22
← cytrinox left17:23
cytrinox_ → cytrinox17:23
→ Shusshu joined17:23
→ kadoban__ joined17:25
← kadoban_ left17:27
← munichlinux left17:27
→ Hoornet joined17:27
← Fandekasp left17:30
→ psoo joined17:32
← Samoi left17:33
rudi_s Is there any documentation about <path>? I'd like to diff all files not having a given extension.17:34
→ _psychic_ joined17:36
← replore left17:36
cbreak try to use git ls-files17:36
and git diff17:37
→ munichlinux joined17:37
cmn a path is used with the usual shell globs, so *.h or **.h would work, but not the other way around17:37
→ bartek joined17:38
→ haseeb joined17:39
→ Octalot joined17:39
→ Targen_ joined17:40
← Targen left17:40
← amigojapan left17:40
← munichlinux left17:42
→ munichlinux joined17:42
rudi_s cbreak: But how can I ignore matches with ls-files?17:43
← mikepack left17:44
→ jharrys joined17:44
→ Jan-Jusky_ joined17:44
cmn grep -v17:46
→ Cromulent joined17:47
rudi_s cmn: Ah, yeah. That works. Thank you.17:48
← rewire left17:50
← dc5ala left17:50
← munichlinux left17:51
← dreiss left17:51
→ munichlinux joined17:51
← jharrys left17:52
→ jharrys joined17:53
← munichlinux left17:55
← cannonball left17:57
rudi_s I'm having some trouble finding a --word-diff-regex which works like vim words. If I have src="img/colortable16.png" and it gets changed to src="http://frexx.de/xterm-256-notes/img/colortable16.png" I'd just like to get the addition of http://frexx.de/... colored. Is this possible?17:58
← _psychic_ left17:58
rudi_s I tried --color-words='[a-zA-Z0-9_]+|[[^[:space:]]+' but that seems to ignore some changes completely.17:59
← mattt_ left17:59
→ mikepack joined17:59
cmn doesn't --word-diff do that?18:01
← Cromulent left18:01
rudi_s cmn: No. It displays src="img/colortable16.png" as removal and src="http://frexx.de/xterm-256-notes/img/colortable16.png" as addition because it separates words by spaces.18:02
→ arvind_khadri joined18:02
← arvind_khadri left18:02
→ arvind_khadri joined18:02
← Hoornet left18:02
← martinjlowm left18:02
rudi_s I'd like words to consist of either a-zA-Z0-9_ or any other characters.18:02
cmn I see18:02
→ Hoornet joined18:03
← mandric left18:04
→ Ben_Shaner joined18:04
← Cogito left18:04
→ macUzer joined18:04
rudi_s Ah, too easy. Looks like --color-words='[a-zA-Z0-9_]+|[^a-zA-Z0-9_]+' works fine.18:05
→ Cogito joined18:05
← Hoornet left18:05
← mikepack left18:06
→ Hoornet joined18:06
← linusoleander left18:08
← Ben_Shaner left18:08
→ mikepack_ joined18:08
→ Ben_Shaner joined18:08
← Ben_Shaner left18:09
→ ben545 joined18:09
← _md left18:09
← lenswipe left18:10
ChanServ set mode: +v18:10
ben545 how do i install git on my server?18:10
kevlarman your package manager18:10
← mikepack_ left18:11
← soc42 left18:13
cbreak from source18:13
jefferai question about internals: when I do a commit, a commit object is created that points to a tree object that itself points to other tree objects/blobs to record the state of the files at that point18:14
← oc80z left18:14
→ oc80z joined18:14
← oc80z left18:14
jefferai what I'm wondering is, when files change18:15
do blobs get created recording the diff of the file, or are blobs created containing the full changed file?18:15
cbreak a blob for a file contains the complete data of the file18:16
afaik git can only deal with single fork files though18:16
→ oc80z joined18:16
jefferai cbreak: so when you commit changes to a file, a blob is created that contains the complete data of the file with your new changes?18:16
← Hoornet left18:16
cbreak only if you commit a new version of that file18:17
it will commit exactly what is staged18:17
jefferai right18:17
so my question then is -- isn't this inefficient?18:17
I know that Git is quite efficient, so I feel like I'm missing something here18:17
cbreak that's where git repack comes into play18:17
→ __CA__ joined18:18
→ Hoornet joined18:20
jefferai cbreak: so I know git repack does delta-compression -- so when repacking, does it take the original blob and then store just the differences between successive changes to that file/later blobs of that file?18:20
cbreak no18:20
SamB cbreak: indeed, git doesn't have a clue "resource forks" (Mac OS), "alternate data streams" (NTFS), and the like18:20
cbreak I think it takes the newest version18:21
SamB jefferai: it just tries to find another blob that's a good basis for compressing18:21
cbreak and then stores deltas to similar blobs, like previous versions18:21
or also other blobs in the same version18:22
jefferai ah, ok18:22
SamB another blob in the same pack18:22
cbreak (for example when you copied files and then modify them a bit)18:22
jefferai right18:22
SamB heck, it wouldn't even have to be a blob18:22
jefferai so it tries to find other blobs that are good basis for compressing, but likely this will be blobs that are previous/later versions of the same blob18:22
SamB it might compress a blob based on a tree or vice-versa18:22
cbreak probably not...18:22
SamB if they were similar enough18:22
jefferai hm, ok18:23
SamB most likely this would only happen if you were storing a git repository in your trees18:23
jefferai so it's not doing anything "smart" aside from more normal delta compression18:23
cbreak I don't think it would happen then18:23
SamB a git repository that was not packed, in fact18:23
cbreak because trees are compressed18:23
→ pheaver joined18:23
← Praise left18:24
→ Praise joined18:25
← Praise left18:25
→ Praise joined18:25
→ Cromulent joined18:25
→ replore joined18:27
jefferai cbreak: SamB: thanks18:27
← ben545 left18:27
→ akostajti joined18:28
← the_metalgamer left18:28
→ xpot-mobile joined18:29
SamB anyway, the real point is that it doesn't really pay attention to semantics at all18:29
← Cromulent left18:30
← Daniel0108 left18:30
← poga left18:31
→ flavius joined18:31
SamB I'm not actually sure it can/would compress objects in terms of objecst of other *types*, but I do know that it pays no attention to the actual history when deciding what objects to compress others in terms of18:32
← w0bni1 left18:32
← __CA__ left18:32
← swombat left18:33
→ Daniel0108 joined18:34
→ swombat joined18:34
cmn that's not quite right, git will try to keep the largest or most recent objects as bases and store the history as deltas of them18:34
← DarkAR left18:35
← jondot left18:35
cmn they tend to be the same according to Linus' law, so working on the most recent files should be just as fast as working with loose objects18:35
→ linusoleander joined18:36
→ jondot joined18:37
← jondot left18:37
joneskoo anyone know if there's going to be improvement regarding gettext po handling some time? from what I understood googling through git mailing list, gettext inclusion is still in progress to git and the last proper discussion I found was this:18:38
http://groups.google.com/group/git-version-control/browse_thread/thread/e5509f268ef5f98b/7e7126dd33c9d034?fwc=118:38
→ d0k_ joined18:39
joneskoo merging po files in a django project is a pain in git. without that problem I could consider a direct-to-production flow after review but now need to do releases to update translations :(18:40
← d0k left18:40
d0k_ → d0k18:40
← swombat left18:40
← dim_ left18:44
← nixmaniack left18:45
← subbyyy left18:45
← jharrys left18:46
→ Cromulent joined18:46
← Cromulent left18:46
→ apok joined18:48
→ soc42 joined18:48
→ _md joined18:49
← dhruvasagar left18:50
→ dhruvasagar joined18:53
→ cjuner joined18:53
→ knifed joined18:54
← arvind_khadri left18:54
→ churp joined18:54
→ berserkr joined18:55
→ pepe joined18:56
← pepe left18:56
← Jan-Jusky_ left18:58
← sythe left18:58
→ infogulch joined18:58
→ jack_rabbit joined18:58
→ arvind_khadri joined19:00
← arvind_khadri left19:00
→ arvind_khadri joined19:00
← BiggFREE left19:01
→ mandric joined19:02
→ BlaDe__ joined19:02
← Textmode left19:02
← BlaDe^ left19:03
← angelsl left19:03
← harry23 left19:04
→ ki__ joined19:04
→ kvanderw joined19:04
→ sythe joined19:07
← sythe left19:07
→ sythe joined19:07
arvind_khadri Hi, I was just wondering, how can I add long commit messages.19:07
jast by writing a long one19:10
→ kvanderw_ joined19:10
jast you should have a lot of space in the editor that launches when you commit stuff19:10
arvind_khadri I do it by git commit -m ""19:11
→ steffo joined19:11
jast that's not very long19:11
arvind_khadri I got it, I just use git commit :)19:11
← kvanderw left19:12
← haseeb left19:15
→ erichynds joined19:16
← ki__ left19:17
← Zelest_ left19:20
cmn arvind_khadri: if you really like -m, you can just keep writing and only put the closing " when you're done writing the message19:22
← JeffJohnson left19:22
arvind_khadri cmn, aah, seems a better way, rather than loading the editor :)19:23
← Tommy[D] left19:23
jast if you never make mistakes and never need to edit your messages, sure19:24
→ Joghurt_ joined19:26
→ Tommy[D] joined19:26
← Joghurt left19:28
← jack_rabbit left19:28
← oc80z left19:30
→ oc80z joined19:30
→ cybersphinx joined19:30
→ pheaver_ joined19:30
→ topping joined19:30
← jimmy1980 left19:32
← pheaver_ left19:33
arvind_khadri jast, but --amend will still work right?19:33
← linusoleander left19:34
← the_cybersphinx left19:34
← pheaver left19:34
→ macmartine joined19:34
→ pheaver joined19:35
→ _psychic_ joined19:35
jast arvind_khadri: sure... but then if you're still not using an editor, you have to completely re-type the message (depending on your shell etc. etc.)19:36
← psoo left19:36
arvind_khadri Oh :)19:36
→ jimmy1980 joined19:37
← sorin left19:39
← pheaver left19:40
→ thierryp joined19:41
→ swombat joined19:41
→ nargin joined19:43
← replore left19:43
← nargin left19:43
← BlaDe__ left19:48
← mandric left19:48
→ BlaDe^ joined19:48
← balau left19:48
→ [1]soc42 joined19:51
→ pablo joined19:52
pablo hi, my projects appear in gitweb as myptoj/.git instead of myproj.git How can I change that?19:53
← _psychic_ left19:53
← kennethreitz left19:54
← soc42 left19:54
→ the_metalgamer joined19:57
jast pablo: you probably want to create them as bare repositories, i.e. inside a foo.git dir, use git init --bare19:58
*automatic message* an explanation of bare and non-bare repositories can be found here: http://bare-vs-nonbare.gitrecipes.de/19:58
← Daniel0108 left19:59
← jelmer_ left20:00
← stamina left20:00
→ kerrick joined20:00
rudi_s Hi. Is it possible to treat carriage return as newline? It screws up diffs with less for me.20:01
→ soc42 joined20:02
→ _psychic_ joined20:02
Corey rudi_s: Yes! http://help.github.com/line-endings/20:03
Well sort of. You convert on commit. :-)20:03
rudi_s Corey: Oh. That's not what I want. I want to keep the file intact (it's a backup of sorts).20:03
jast their tips are... not so great20:03
→ mandric joined20:04
jast rudi_s: git diff -b/-w20:04
→ the_cybersphinx joined20:04
← arvind_khadri left20:04
← [1]soc42 left20:05
→ dreiss joined20:06
→ cirwin joined20:06
rudi_s jast: Thanks, this works most of the time. Not 100% in this case. I guess there's no, pass all files to tr '\r' '\n' before diffing? Or should I use textconv for that?20:07
→ fbdystang joined20:07
← cybersphinx left20:08
← erichynds left20:09
→ Daniel0108 joined20:09
← thierryp left20:15
→ jharrys joined20:16
→ arvind_khadri joined20:16
← jharrys left20:18
jast rudi_s: probably20:18
← omf_ left20:18
→ eean joined20:19
← LongBeach left20:19
→ jelmer_ joined20:21
rudi_s jast: Ok, thanks. But how can I specify tr '\r' '\n' in my .gitconfig. I tried textconv = tr '\r' '\n' but get the error "bad config file line 106"20:22
jast rudi_s: see the gitattributes manpage. you may have to make it call a script rather than run tr with arguments directly.20:24
cmn textconv isn't a script; you might want to take a look at file attributes and the crlf conversion20:24
jast gotta go now. good luck.20:24
rudi_s jast: Thanks for your help.20:24
cmn: I know, it's a program to run. The quotes seem to confuse git, but why?20:25
jast textconv settings don't belong in .gitconfig20:25
rudi_s ?20:25
jast as I said20:25
read the gitattributes manpage20:25
which is not the git-config manpage ;)20:25
and now, good night20:25
rudi_s Good night.20:25
22:25 < jast> textconv settings don't belong in .gitconfig20:26
Why not? man gitattributes even states that: "(or $HOME/.gitconfig file):"20:26
jast *automatic message* the 'gitattributes' manpage can be found at http://www.kernel.org/pub/software/scm/git/docs/gitattributes.html20:26
→ murilobr joined20:26
→ kadoban joined20:26
← kadoban__ left20:27
→ jharrys joined20:29
rudi_s Ok .. should've read man tr .. it doesn't take a filename. Helper script it is then.20:30
→ poseid joined20:30
← Daniel0108 left20:31
→ Daniel0108 joined20:31
← kadoban left20:36
← j416 left20:36
→ Drakonite2 joined20:36
→ j416 joined20:36
← Drakonite left20:37
Drakonite2 → Drakonite20:37
→ metoo_ joined20:40
→ subbyyy joined20:40
← the_metalgamer left20:40
← dSebastien left20:40
← werdan7 left20:44
← unnamedwill left20:44
← k6b left20:44
→ unnamedwill joined20:45
→ [1]soc42 joined20:45
← candybar left20:45
→ k6b joined20:45
← ezkl left20:46
→ DarkAR joined20:46
k6b → Guest7217520:46
→ candybar joined20:46
→ pantsman joined20:46
← pantsman left20:46
→ pantsman joined20:46
← rdz left20:47
← coldboot left20:47
→ coldboot joined20:48
← soc42 left20:48
← xpot-mobile left20:48
← metoo_ left20:48
→ Pacman_Home joined20:48
→ ezkl joined20:49
→ rdz joined20:49
→ SimonNa joined20:50
→ xpot-mobile joined20:50
→ jehoshua02 joined20:51
← Pacman left20:51
→ dr_lepper joined20:51
→ soc42 joined20:52
jehoshua02 There's a couple git ftp tools. Has anybody used them?20:52
SethRobertson For pushing to a web site that doesn't support git or ssh or rsync?20:52
→ kadoban joined20:53
← glennpratt left20:53
→ Joefish joined20:53
← Joefish left20:53
→ johnkpaul joined20:53
jehoshua02 SethRobertson: Yes, yes, yes. I'm jumping in for emergency help, and the team that's been working on this has no version control.20:54
→ flamingspinach joined20:54
→ MrKeuner joined20:55
← MrKeuner left20:55
→ MrKeuner joined20:55
← [1]soc42 left20:55
SethRobertson I have heard of them but not used them. I strongly suggest using rsync if git is not available on the web server. I don't specifically have a recommendation for the git-ftp tools. Someone asked about them once but it turned out he could use rsync after all.20:55
→ jasonmc joined20:56
← jasonmc left20:56
jehoshua02 SethRobertson: All they've given me is ftp access.20:56
← dangerousdave left20:57
MrKeuner hello, if I remove a code snippet from lines 20-29 of project.java then after several changes to the project.java where does the 10 line code go back if I revert the change of removing code snippet? to the 20-29 again?20:57
SethRobertson jehoshua02: Such is life.20:57
→ nikhgupta joined20:57
cirwin MrKeuner: to the place in the file that "looks right" to git20:57
MrKeuner how does git decide?20:58
SethRobertson MrKeuner: It will go to lines 20-29 +- some fuzz factor, by matching the lines surrounding your context20:58
cirwin so it will look at the preceding and subsequent few lines to try and find where the lines used to be20:58
jehoshua02 SethRobertson: Git over ftp is a good idea no? As long as it behaves just like push.20:58
→ nimrod10` joined20:58
SethRobertson MrKeuner: If you change the lines around the lines you deleted, you will get a conflict and git will ask you to figure it out manually.20:58
MrKeuner i see20:59
SethRobertson jehoshua02: Your situation is NOT git over ftp. It is web site publishing over ftp. However, this can be triggered via a `git push` if you want.20:59
← subbyyy left20:59
MrKeuner thanks20:59
→ AaronMT joined21:00
jehoshua02 SethRobertson: Are you saying that git push can use ftp?21:01
← nimrod10 left21:01
← MrKeuner left21:01
SethRobertson jehoshua02: Not at all. What you can have is a repo which has a post-receive trigger which calls a of the website publishing script which uses git. So when you are ready to publish you can `git push live` which sends over to the live repo which then automatically runs the script to ftp everything over to the ftp server.21:02
I don't specifically know if ftp works to allow pushes to a git repo or not. I don't remember seeing that change. It certainly didn't use to work.21:04
But as I said, I don't think that is your use case anyway21:04
← jharrys left21:04
jehoshua02 SethRobertson: So this team, they aren't using version control, and I don't want to influence them to do so. So there will be no central repo . . .21:05
← LouisJB left21:05
SethRobertson Just because *they* don't have a central repo doesn't mean that *you* cannot have one21:05
jehoshua02 SethRobertson: Exactly. I'm just wondering how to make sure my changes are deployed with out overwritting their changes, and vice-versa, when all I have is ftp.21:06
← apok left21:07
← baburdick left21:09
cirwin jehoshua02: you can't remove all the race conditions, but if you always download the latest version from ftp before pushing, you're "unlikely" to overwrite their changes21:09
→ baburdick joined21:10
SethRobertson jehoshua02: That is of course impossible, but you can give it your best shot, subject to race conditions. If I were in such a undesirable situation I would get an ftp-mirror script and a ftp-publish script. I would set things up so that when I ran the ftp-mirror script it would check out a "vendor" branch, take all changes, then commit them. I would then merge the vendor branch into my master branch. work work work21:10
test test test. Commit. ftp-mirror again. If there were changes, back to testing. If no changes, then publish via ftp. Then cause the vendor branch to be overwritten with the current changes on master.21:10
Guest72175 → k6b21:10
→ jonanin joined21:11
→ Ig0r_ joined21:11
jehoshua02 SethRobertson: One challenge is removing files. If they remove a file, I will still have the file after downloading. Or if I remove a file, it won't be removed from live.21:14
← johnkpaul left21:14
SethRobertson I think the git-ftp publishing tools are aware of this.21:14
← kerrick left21:15
SethRobertson For publishing. Same with the ftp-mirror script. You can nuke everything in the git repo, then run mirror, then run git add -A, and git will know about the deletions21:15
← _md left21:16
jehoshua02 SethRobertson: I thought of that, but this is a decent sized project, so nuking it every time could be time consuming... Hmmm....21:16
→ stamina joined21:17
→ jerkku joined21:17
SethRobertson Nuking is fast. Re-mirroring is slow.21:17
jehoshua02 SethRobertson: Right.21:17
SethRobertson: I really think a git-ftp solution is a good idea, I'm just wondering how "aware" these ftp tools are.21:18
← poseid left21:18
SethRobertson If you want to have a clean mirror and your only tool is ftp, you don't have many choices. Try asking for rsync access, or perhaps tarballs of the website21:18
As I said, I have never used them. However, since they have all of the information available to them, I would be surprised if they did not support deletion21:19
jehoshua02 SethRobertson: I'd also hope these git-ftp tools prevent uploads if there have been changes, like push, requiring you to stop and pull down new stuff...21:20
SethRobertson: I guess I have no choice but to try!21:20
SethRobertson The probably CANNOT do that. Because ftp is too stupid of a protocol.21:21
That is why I suggested the workflow I did21:21
jehoshua02 SethRobertson: It can't be that stupid... afterall, my ftp client (filezilla) has directory comparison. And perhaps these git-ftp tools use that kind of workflow.21:22
← brendon left21:23
← cyphase left21:23
jehoshua02 SethRobertson: I'll try these tools out and see what happens.21:23
← thiago left21:24
→ cyphase joined21:24
Ig0r_ Hi everyone! Can someone please point me to documentation about this question:21:25
→ Joefish joined21:25
Ig0r_ I have local dev files. And I have a server with those files (I've uploaded them earlier by ftp)21:25
How can I synchronize them now? So that when I commit and push to that server from my local machine - files are updated on server too?21:25
← Raging_Hog left21:26
→ jatt joined21:26
Ig0r_ All docs that I've found - give examples of pushing to git --bare repository, which doesn't have a workout directory and no files are updated(21:26
→ jack_rabbit joined21:26
← nikhgupta left21:26
→ axl_ joined21:27
→ nikhgupta joined21:27
→ poseid joined21:27
cmn you don't want to do that; you might want to take a look at web deployment using git21:27
j416 Ig0r_: git isn't the best deployment tool.21:28
SethRobertson Ig0r_: Do you have access to the web server by some method other than ftp?21:28
cmn there are many turorials about that, but the main point is to use GIT_WORKTREE=/var/www/whatever git checkout -f in the post-push21:28
post-receive hook, I mean21:28
Ig0r_ sure ssh21:28
j416 Ig0r_: you can accomplish it using hooks and that, but imo it's more hassle than it's worth.21:28
Ig0r_: rsync is a simple solution..21:29
← bindaasomatic left21:30
SethRobertson Ig0r_: http://toroid.org/ams/git-website-howto http://stackoverflow.com/questions/6112852/git-website-update-strategy-how-to-sync-dev-and-live-repositories21:30
Ig0r_ Interesting, so you suggest to look at deployment tools like Capistrano etc?21:31
→ chrisf__ joined21:31
← soc42 left21:31
→ talios joined21:31
SethRobertson If the Capistrano framework works for you, yes21:32
→ the_metalgamer joined21:32
SethRobertson git can fake a web deployment system for a sufficiently simple website, but tools written for the job are better21:32
← nikhgupta left21:33
→ celestius joined21:33
Ig0r_ Great info! Thank you very much guys!21:35
→ caseymcg joined21:36
→ haseeb joined21:38
← tewecske left21:38
← Joghurt_ left21:40
→ Joghurt_ joined21:40
← poseid left21:41
← alip left21:42
→ jzacsh-shore joined21:43
→ alip joined21:45
← Hoornet left21:46
← jack_rabbit left21:48
← jzacsh-shore left21:48
← cyphase left21:49
→ poseid joined21:49
k6b → k6b`21:52
k6b` → k6b21:52
→ zacsh joined21:53
zacsh hi, i have a directory (~/.vim/bundle/) which has git-submodule's `add`ed to it... for some reason when a submodule (which happens to be a plugin) adds a file there (eg.: Command-T plugin created: ~/.vim/bundle/Command-T/.netrwhist), sees that like so: http://dpaste.de/Di1cW/raw/21:53
anyone know what i can do to solve the issue and ignore changes to submodules in .vim/bundle/ ? I added .vim/bundle/* to my repo's .gitignore, but it didn't help...21:54
← afief left21:54
← axl_ left21:54
jakobm zacsh: as indicated by the message, those changes will not be commited21:56
you just want them not to be displayed?21:56
→ boombatower joined21:58
jehoshua02 I want to ignore everything in "admin/uploads/" but it doesn't seem to be working.21:58
← haseeb left21:58
zacsh jakobm: yeah, i mean... i'd like a clean working directory. i'd imagine that message is indication that things aren't clean and should be attended to21:58
jehoshua02 git status is still showing all changes made in "admin/uploads/"21:58
→ nicoulaj joined21:59
→ camb joined21:59
SethRobertson zacsh: Perhaps gitslave would work better for you than git-submodules21:59
← kenperkins left22:00
zacsh googles gitslave22:00
SethRobertson zacsh: However, in general any file which has been added to git cannot be ignored WRT changes. And no. assumed unchanged is not a good idea.22:00
gitslave.sf.net22:00
← Shusshu left22:01
zacsh SethRobertson: what do you mean "assumed unchanged"?22:01
SethRobertson google git assumed unchanged, and then don't use it22:02
→ CannedCorn joined22:02
zacsh SethRobertson: ah. i see. well. ... i like git telling me when things change..22:03
i just don't want it to watch that folder. but, perhaps you're right and this just means i shouldn't be using submodule for the task..22:03
← talios left22:04
← xpot-mobile left22:04
← eeanm left22:04
← knifed left22:04
← cirwin left22:04
jehoshua02 When I put "admin/upload/" in .gitignore, should I still see files in that directory in "git status" output?22:05
→ cirwin joined22:05
← jerkku left22:05
SethRobertson jehoshua02: If those files were committed to git, yes22:05
It seems like everyone is asking the same question here today, right after each other.22:06
jehoshua02 SethRobertson: Hmm...22:06
→ kadoban_ joined22:07
← kadoban left22:08
← poseid left22:08
zacsh SethRobertson: thanks!22:09
jakobm: ^ thanks22:09
zacsh waves to room22:09
← Cogito left22:09
← tatsuya_o left22:10
jehoshua02 SethRobertson: Well that's confusing... I guess git has a different idea of ignore than humans do...22:10
→ kadoban__ joined22:10
→ Cogito joined22:10
← berserkr left22:10
jehoshua02 I would have thought "ignore" meant "no touchy".22:10
SethRobertson .gitignore is ONLY for untracked files.22:10
→ OrionX joined22:11
cmn yes, unless you've explicitely told git to take care of it22:12
which is what you do when you start tracking a file22:12
← kadoban_ left22:12
← ivorybishop left22:13
jehoshua02 So then, is there a way to untrack a file without modifying history?22:13
SethRobertson git [--cached] rm22:13
→ khmarbaise joined22:13
← Orion_X left22:13
jehoshua02 consulting 'git help git'22:13
SethRobertson man git-rm22:13
jast *automatic message* the 'git-rm' manpage can be found at http://www.kernel.org/pub/software/scm/git/docs/git-rm.html22:13
SethRobertson Well, you can't see it at that url at the moment22:14
jehoshua02 jast: kernel.org is down.22:14
← zacsh left22:14
← camb left22:14
jehoshua02 Okay. I'll look at that.22:14
Thanks.22:15
← sythe left22:15
← mandric left22:17
← aidenhong left22:18
→ mandric joined22:18
← khmarbaise left22:19
← Joghurt_ left22:20
→ berserkr joined22:23
← bolt left22:24
← steffo left22:25
← niloc132 left22:26
← nicoulaj left22:26
→ bolt joined22:27
← Spockz left22:27
→ Sajaki joined22:28
← f0i left22:28
→ niloc132 joined22:28
→ f0i joined22:29
→ Fuwex joined22:31
← pantsman left22:31
→ Samoi joined22:32
← Samoi left22:32
→ Samoi joined22:32
← j416 left22:35
→ tshauck joined22:36
→ EricInBNE joined22:37
→ kerrick joined22:38
→ wbednarski joined22:38
← Ig0r_ left22:40
← murilobr left22:40
→ shiba_yu36 joined22:43
← fisted left22:45
← the_metalgamer left22:47
← Amorphous left22:49
→ jmah joined22:51
← sp3001 left22:53
→ Fandekasp joined22:55
← DarkAR left22:57
← d0k left22:57
← mandric left22:58
pablo jast: thank you very much22:58
← akostajti left22:58
pablo jast: that was the solution22:58
jast: git init --bare22:58
jast: thanks22:58
← pablo left22:58
→ tdebat joined23:00
→ jack_rabbit joined23:01
← ISF left23:01
← whitman left23:02
← cheater left23:02
→ icwiener joined23:03
← Daniel0108 left23:05
→ Amorphous joined23:06
→ psoo joined23:07
→ krazyj joined23:08
krazyj hey folks… git noob here. i just did a `git stash pop` and have some conflicts. i just want to resolve it by going with the file, as it is in the stash.23:09
any ideas?23:09
cmn does checkout --theirs work?23:10
krazyj i'm using Tower for my GUI interactions and tried using a 'resolve by using ours' functionality - not sure what that did but it *didnt* fix it23:10
$ git checkout --theirs23:10
fatal: --ours/--theirs is incompatible with switching branches.23:10
← seed419 left23:11
cmn you need to tell it the file23:11
krazyj cmn: works23:12
could you explain what i did there?23:12
checkout --theirs, *doesn't* move _their_ copy into my local branch?23:12
cmn you told git to checkout the file using whatever version "they" had23:12
← jatt left23:13
→ bobdobbs joined23:13
cmn it's not about branches23:13
krazyj ah k23:13
ill have to look into that23:13
thanks cmn23:13
cmn it's about resolving conficts (in a merge, it is from another branch)23:14
→ Dave^|| joined23:14
krazyj in my situation, what should be done for a file that is 'added by us' versus a 'both modified' file?23:14
← jakobm left23:14
← superdmp left23:15
→ cher joined23:15
cmn that depends on what you want to do, for both modified, it's --ours or --theris23:15
s/theris/theirs/23:15
for addede/removed files, add or rm23:16
→ sythe joined23:16
krazyj ah ok23:17
i did checkout --theirs on all the conflicted files23:17
however a git status still lists them as unmerged23:17
even though it looks like they were replaced?23:17
→ vincentmac joined23:17
← Dave^| left23:17
cmn stage them23:17
krazyj by golly, that's done it23:18
thanks for your help cmn23:19
← macmartine left23:20
← krazyj left23:21
→ Daniel0108 joined23:21
← boombatower left23:21
← SimonNa left23:22
→ berserkr1 joined23:23
← arvind_khadri left23:24
← CannedCorn left23:24
← kvanderw_ left23:25
← psoo left23:25
← berserkr left23:27
jehoshua02 Do I have to track .gitignore?23:30
← ch077179 left23:30
jehoshua02 Will .gitignore still take effect if not tracked?23:31
SethRobertson no myes23:31
jmah I think so23:31
→ kvanderw joined23:31
SethRobertson no, yes23:31
m1sc jehoshua02: you know about .git/info/exclude ?23:31
→ reecefowell joined23:31
jmah you can add a user-local ignore with core.excludesfile instead23:32
(or info/exclude, yes)23:32
→ Chib joined23:32
← Dave^|| left23:32
SethRobertson All three are options23:33
← kerrick left23:33
reecefowell how to i move my git project to start from another directory? i have this so far https://github.com/reecefowell/CCDNForum but i want to loose the vendors dir and make the project start from inside the src folder without showing the folder name, so just show the folder CodeConsortium23:34
so show /src/codeconsortium contents instead of the contents of /23:35
jmah reecefowell: look up git-filter-branch, the subtree-filter option23:35
--subdirectory-filter, rather23:35
← Daniel0108 left23:37
← _psychic_ left23:38
reecefowell jmah: but that will create a new repo?23:39
i want to keep the one i got23:39
→ Spawnrad joined23:40
jmah no (you can always join / split repos anyway)23:40
reecefowell but i have to clone?23:41
jmah no...23:42
where did it say that?23:42
reecefowell here http://www.nofluffjuststuff.com/blog/matthew_mccullough/2010/12/git_filter_branch_examples23:43
in the pdf23:43
all i want to do is make the repo start from the subdirectory instead of the parent directory23:44
jmah you'd usually make a clone just to have a backup23:44
← berserkr1 left23:44
→ BiggFREE joined23:44
jmah but there's no requirement to23:44
reecefowell aye i don't wish to make a backup23:45
jmah in the man page for git-filter-branch, look up "foodir"23:45
there's an example there23:46
→ ch077179 joined23:46
← wbednarski left23:46
reecefowell would be easier to just delete the repo and make a new one at this point23:46
SethRobertson Whatever makes you happy. it isn't necessary23:47
→ g0rdorin joined23:47
→ lightcap joined23:48
← cher left23:48
← babilen left23:48
← Vampire0 left23:48
→ babilen joined23:48
→ Vampire0 joined23:49
← Fuwex left23:49
→ Fuwex joined23:50
reecefowell jmah: i tried it but i get this23:53
eece-Fowells-Mac-Pro:symfony reecefowell$ git filter-branch --subdirectory-filter src/CodeConsortium/ -- --all23:53
Cannot rewrite branch(es) with a dirty working directory.23:53
jmah do you know what "dirty working directory" means?23:53
reecefowell something must be in there it does not like i guess23:54
honestly though no idea23:54
jmah there are uncommitted changes23:55
so either commit them or revert the files23:55
reecefowell oh i see23:55
ok thats done but now i cannot push it to github23:57
← Vampire0 left23:57
← g0rdorin left23:58
reecefowell http://pastie.org/private/bpoitxugpeux6ockxo0w23:58
← Samoi left23:58
→ Samoi joined23:58
← fbdystang left23:58
jmah push with -f23:59
because you re-wrote history23:59

Logs Search ←Prev date Next date→ Channels Documentation